Developed by TrackMan, the professional baseball‘s first AI umpire comes in the form of an automated strike zone.
As reported by ESPN, umpire Brian deBrauwere had an iPhone with him that was connected to the TrackMan system. And through an AirPod headphone, deBrauwere heard the system's calls, and signaled balls and strikes as he normally would.
The plan to try this technology was first announced back in March 2019, and after a half-season of testing, the radar-powered system was deemed ready to go for the All-Star game.
It's an experiment done in partnership with Major League Baseball.
The Rockers and the Atlantic League are an independent baseball league not directly affiliated with Major League Baseball (MLB), but the AI systems in use are part of a partnership between the two.
In a three-year deal, the Atlantic League’s players and coaches are testing this technology for MLB, in order to see if automated systems are right for prime time sports.
During the game, the strike zone was different than they were used to, because the TrackMan system is better at discerning certain pitches than humans.
"The consensus among players and umpires who have tested it is that TrackMan squeezes the corners of the plate where human umpires might not, and grants strikes higher and lower in the zone," wrote The Washington Post explains.
Initially, the data collected from TrackMan is used in a wide variety of scenarios, but in-game pitch calls aren't one of them.
During the game, the AI did made some mistakes, like when it robbed the Rockers of an earned strikeout.
And here, Frank Viola disagreed with a call the TrackMan made, but the human umpire upheld it.
After a heated argument, Viola became the first person thrown from a professional baseball game for arguing with a human umpire over the call the AI made.
The AI used in the game was its first version, and during TrackMan AI‘s first day on the job, the players and coaches haven’t responded as positively to the system as its developers might have predicted.
But still, the technology should become better in time. Eventually, systems like these should become more robust.
It has long been problematic for players to rely on a centuries-old system built on old-fashioned human eyeballing.
Given that modern computers has the ability beyond humans in determining a fact in a split of a second, the idea of using AI is seen a a great idea.
The goal of this technology is to aid human umpire, as well as allowing everyone who watches the game from home to see whether a pitch is called a ball or a strike immediately.
This can be made possible using the tracking system's data being used in game broadcasts.
